The present invention relates to an image processing apparatus for detecting a specific original.
In recent years, along with the improvements in the performance of color copying machines and color printers, illegal uses of these apparatuses pose a problem. More specifically, these apparatuses are used for counterfeiting banknotes (paper money), securities, and the like. For this reason, techniques for preventing copies of specific originals are demanded.
As one of such techniques, the present applicant has proposed the following technique (U.S. Pat. No. 5,227,871). That is, the distribution information of color tones, on a color space, of each specific original is registered in advance, and whether or not input image data is one of the registered specific originals is discriminated by comparing the registered distribution information with the color tone distribution of the input image data.
However, the above-mentioned conventional specific original discrimination method suffers from the problem that the number of types of specific originals to be discriminated is very large, i.e., the number of color tone distribution information to be registered in advance in a color copying machine or a color printer is very large. If only banknotes are to be discriminated, the number of types of banknotes to be registered exceeds several tens in consideration of differences in designs on their two sides, and necessity of registration of banknotes issued in surrounding countries in, especially, Europe.
In order to improve the discrimination rate of a specific original, a single specific original is preferably discriminated using a plurality of types of tone color distribution information. Although the discrimination conditions become stricter, the number of color tone discrimination information to be used in discrimination is preferably as large as possible to eliminate discrimination errors.